Search NASA⌕ Search

SEARCH · Search NASA

Results for “Topological optimization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 523 records · Page 29

Computations of Unsteady Viscous Compressible Flows Using Adaptive Mesh Refinement in Curvilinear Body-fitted Grid Systems

A methodology for accurate and efficient simulation of unsteady, compressible flows is presented. The cornerstones of the methodology are a special discretization of the Navier-Stokes equations on structured body-fitted grid systems and an efficient solution-adaptive mesh refinement technique for structured grids. The discretization employs an explicit multidimensional upwind scheme for the inviscid fluxes and an implicit treatment of the viscous terms. The mesh refinement technique is based on the AMR algorithm of Berger and Colella. In this approach, cells on each level of refinement are organized into a small number of topologically rectangular blocks, each containing several thousand cells. The small number of blocks leads to small overhead in managing data, while their size and regular topology means that a high degree of optimization can be achieved on computers with vector processors.

Steinthorsson, E.↗

ARPA-E: Advancing the Electric Grid

The electric grid was designed with the assumption that all energy generation sources would be relatively controllable, and grid operators would always be able to predict when and where those sources would be located. With the addition of renewable energy sources like wind and solar, which can be installed faster than traditional generation technologies, this is no longer the case. Furthermore, the fact that renewable energy sources are imperfectly predictable means that the grid has to adapt in real-time to changing patterns of power flow. We need a dynamic grid that is far more flexible. This video highlights three ARPA-E-funded approaches to improving the grid's flexibility: topology control software from Boston University that optimizes power flow, gas tube switches from General Electric that provide efficient power conversion, and flow batteries from Harvard University that offer grid-scale energy storage.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Relations between environmental noise and electronic coupling for optimal exciton transfer in one- and two-dimensional homogeneous and inhomogeneous quantum systems

Recent studies have indicated that environmental noise may increase energy-transfer efficiency in quantum systems. For homogeneous networks of chromophores previous studies have primarily considered excitonic transport in one-dimensional (linear) networks. In our study, we expand previous research to a two-dimensional fully coupled topology of chromophore molecules. We demonstrate that not only does an optimal dephasing rate exist in both one- and two-dimensional networks but also that it increases in magnitude with increasing coupling strength between chromophores. Optimal transport occurs when the noise quenches the entanglement between local modes that prevent the exciton from moving efficiently to the target site. We find that these results are insensitive to minor site defects such as those found in realistic systems. We contrast these findings to systems with a high degree of inhomogeneity, in which the optimal dephasing rate is largely set by the system topology and does not vary significantly with respect to coupling strength. Our findings have potential applications to systems such as quantum dot arrays and carbon nanotube structures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Topological superfluid in a trapped two-dimensional polarized Fermi gas with spin-orbit coupling

We study the stability region of the topological superfluid phase in a trapped two-dimensional polarized Fermi gas with spin-orbit coupling and across a BCS-BEC crossover. Due to the competition between polarization, pairing interaction, and spin-orbit coupling, the Fermi gas typically phase-separates in the trap. Employing a mean-field approach that guarantees the ground-state solution, we systematically study the structure of the phase separation and investigate in detail the optimal parameter region for the preparation of the topologically nontrivial superfluid phase. We then calculate the momentum space density distribution of the topological superfluid state and demonstrate that the existence of the phase leaves a unique signature in the trap integrated momentum space density distribution which can survive the time-of-flight imaging process.

74 ATOMIC AND MOLECULAR PHYSICS↗

An analytical approach to evaluating magnetic-field closure and topological changes in FRC devices

Here, we describe mathematical methods based on optimizing a modified non-linear flux function (MFF) to evaluate whether odd-parity perturbations affect the local closure of magnetic field lines in field-reversed configurations. Using the MFF methodology, quantitative formulas are derived that provide the shift of the field minimum and the threshold for field-line opening, a discontinuous change in field topology.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Aerodynamic Shape Optimization Based on Free-form Deformation

This paper presents a free-form deformation technique suitable for aerodynamic shape optimization. Because the proposed technique is independent of grid topology, we can treat structured and unstructured computational fluid dynamics grids in the same manner. The proposed technique is an alternative shape parameterization technique to a trivariate volume technique. It retains the flexibility and freedom of trivariate volumes for CFD shape optimization, but it uses a bivariate surface representation. This reduces the number of design variables by an order of magnitude, and it provides much better control for surface shape changes. The proposed technique is simple, compact, and efficient. The analytical sensitivity derivatives are independent of the design variables and are easily computed for use in a gradient-based optimization. The paper includes the complete formulation and aerodynamics shape optimization results.

Samareh, Jamshid A.↗

Laser-driven direct quantum control of nuclear excitations

The possibility of controlled direct laser-nuclear excitations is considered from a quantum control perspective. The controllability of laser-driven electric dipole and magnetic dipole transitions among pure nuclear states is analyzed. Within a set of realistic and general conditions, atomic nuclei are demonstrated to possess full state controllability. Additionally, an analysis of the nuclear state excitation probability as a function of the laser control field is conducted. This control landscape is shown to possess a generic topology, which has important physical consequences for achieving optimal nuclear state excitation with laser fields. Last, an assessment is given of the technological challenges that need to be considered when implementing direct nuclear control in the laboratory.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Progress Toward Attractive Stellarators

The quasi-axisymmetric stellarator (QAS) concept offers a promising path to a more compact stellarator reactor, closer in linear dimensions to tokamak reactors than previous stellarator designs. Concept improvements are needed, however, to make it more maintainable and more compatible with high plant availability. Using the ARIES-CS design as a starting point, compact stellarator designs with improved maintenance characteristics have been developed. While the ARIES-CS features a through-the-port maintenance scheme, we have investigated configuration changes to enable a sector-maintenance approach, as envisioned for example in ARIES AT. Three approaches are reported. The first is to make tradeoffs within the QAS design space, giving greater emphasis to maintainability criteria. The second approach is to improve the optimization tools to more accurately and efficiently target the physics properties of importance. The third is to employ a hybrid coil topology, so that the plasma shaping functions of the main coils are shared more optimally, either with passive conductors made of high-temperature superconductor or with local compensation coils, allowing the main coils to become simpler. Optimization tools are being improved to test these approaches.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Speed Trajectory Optimization for a Heavy-Duty Truck Traversing Multiple Signalized Intersections: A Dynamic Programming Study

This paper explores the fuel savings that can be achieved by optimizing the speed trajectory of a heavy-duty truck traversing a sequence of intersections, under the assumptions that the behavior of the leading traffic and the timing of the traffic lights is known. Specifically, we look at the impact of corridor topology (i.e. green cycle lengths, phase offsets) on the expected fuel savings of the optimized trajectories. This is an important area of research because vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) technology has the potential to allow autonomous vehicles to reduce fuel consumption, especially in urban and sub-urban driving scenarios. The literature tackles the problem of arterial corridor trajectory optimization, and shows the potential fuel saving benefits. However, previous research focuses primarily on passenger vehicles, and often limits its findings to specific case studies. The main contribution of this paper is to offer an estimate of the fuel saving potential - for heavy-duty trucks and under different corridor characteristics - of optimizing trajectories in an urban arterial with V2V and V21 capabilities.

42 ENGINEERING↗

TGE: Machine Learning Based Task Graph Embedding for Large-Scale Topology Mapping

Task mapping is an important problem in parallel and distributed computing. The goal in task mapping is to find an optimal layout of the processes of an application (or a task) onto a given network topology. We target this problem in the context of staging applications. A staging application consists of two or more parallel applications (also referred to as staging tasks) which run concurrently and exchange data over the course of computation. Task mapping becomes a more challenging problem in staging applications, because not only data is exchanged between the staging tasks, but also the processes of a staging task may exchange data with each other. We propose a novel method, called Task Graph Embedding (TGE), that harnesses the observable graph structures of parallel applications and network topologies. TGE employs a machine learning based algorithm to find the best representation of a graph, called an embedding, onto a space in which the task-to-processor mapping problem can be solved. We evaluate and demonstrate the effectiveness of TGE experimentally with the communication patterns extracted from runs of XGC, a large-scale fusion simulation code, on Titan.

Choi, Jong Youl↗

Use of density functional theory method to calculate structures of neutral carbon clusters C{sub n} (3 ≤ n ≤ 24) and study their variability of structural forms

In this work, we present modifications to the well-known basin hopping (BH) optimization algorithm [D. J. Wales and J. P. Doye, J. Phys. Chem. A 101, 5111 (1997)] by incorporating in it the unique and specific nature of interactions among valence electrons and ions in carbon atoms through calculating the cluster’s total energy by the density functional tight-binding (DFTB) theory, using it to find the lowest energy structures of carbon clusters and, from these optimized atomic and electronic structures, studying their varied forms of topological transitions, which include a linear chain, a monocyclic to a polycyclic ring, and a fullerene/cage-like geometry. In this modified BH (MBH) algorithm, we define a spatial volume within which the cluster’s lowest energy structure is to be searched, and introduce in addition a cut-and-splice genetic operator to increase the searching performance of the energy minimum than the original BH technique. The present MBH/DFTB algorithm is, therefore, characteristically distinguishable from the original BH technique commonly applied to nonmetallic and metallic clusters, technically more thorough and natural in describing the intricate couplings between valence electrons and ions in a carbon cluster, and thus theoretically sound in putting these two charged components on an equal footing. The proposed modified minimization algorithm should be more appropriate, accurate, and precise in the description of a carbon cluster. We evaluate the present algorithm, its energy-minimum searching in particular, by its optimization robustness. Specifically, we first check the MBH/DFTB technique for two representative carbon clusters of larger size, i.e., C{sub 60} and C{sub 72} against the popular cut-and-splice approach [D. M. Deaven and K. M. Ho, Phys. Rev. Lett. 75, 288 (1995)] that normally is combined with the genetic algorithm method for finding the cluster’s energy minimum, before employing it to investigate carbon clusters in the size range C{sub 3}-C{sub 24} studying their topological transitions. An effort was also made to compare our MBH/DFTB and its re-optimized results carried out by full density functional theory (DFT) calculations with some early DFT-based studies.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

SMALE: Enhancing Scalability of Machine Learning Algorithms on Extreme-Scale Computing Platforms

Deployment and execution of machine learning tasks on extreme-scale computing platforms face several significant technical challenges: 1) High computing cost incurred by dense networks – The computing workload of deep networks with densely-connected topology increases rapidly with the network size, imposing a non-scalable computing model of extreme-scale computing platforms; 2) Non-optimized workload distribution – Many advanced deep learning algorithms, e.g., sparsification and irregular net-work topology, produce very unbalanced workload distribution on extreme-scale computing platforms. The computation efficiency is greatly hindered by the incurred data and computation redundancies as well as long tails of the node with extensive workload; 3) Constraints in data movement and I/O bottle-neck – Inter-node data movement in extreme-scale computing platforms are associated with high energy and latency costs, and subject to the constraints of I/O bandwidth; and 4) Generalization of algorithm realization and acceleration on computing platforms – The large varieties of machine learning algorithms and structures of extreme-scale computing platforms make the derivation of a generalized algorithm realization and acceleration method very challenging, which, however, is the requirement by domain scientists and interested users. We call the above challenges Smale’s Problems in Machine Learning and Understanding for High-Performance Computing Scientific Discovery. The objective of our three-year research project is to develop a holistic innovation set at structure, assembly, and acceleration layers of machine learning algorithms to address the above challenges in algorithm deployment and execution. Three tasks are particularly performed, including: At the algorithm structure level, we investigate the techniques that can structurally sparsify on the topology of deep networks for computing workload reduction. We also study clustering and pruning techniques that can optimize the workload distributions over the extreme-scale computing platforms; At the algorithm assembly level, we derive a unified learning framework for unsupervised transfer learning and dynamic growing capabilities. Novel training methods are also exploited to enhance the training efficiency of the proposed framework; At the algorithm acceleration level, we will develop a series of techniques that can accelerate the computation of sparse matrix operations, which are one of the core executions in deep learning and optimize memory access of the concerned platforms. Our proposed techniques attack the fundamental problems in machine learning algorithms running on extreme-scale computing platforms by vertically integrating the solutions at three closely entangled layers, paving the long-term scaling path of machine learning applications under DOE context. Three tasks corresponding to the above respective research orientations are performed during the three-year project period with our collaborators at ORNL. The outcome of the proposed project is anticipated to form a holistic solution set of novel algorithms and network topologies, efficient training techniques, and fast acceleration methods to promote the computing scalability of the machine learning applications of particular interest to DOE.

97 MATHEMATICS AND COMPUTING↗

Compact toroidal ion-trap design and optimization

We present the design of a type of compact toroidal, or 'halo', ion trap. Such traps may be useful for mass spectrometry, studying small Coulomb cluster rings, quantum-information applications, or other quantum simulations where a ring topology is of interest. We present results from a Monte Carlo optimization of the trap design parameters using finite-element analysis simulations that minimize higher-order anharmonic terms in the trapping pseudopotential, while maintaining complete control over ion placement at the pseudopotential node in three dimensions using static bias fields. These simulations are based on a practical electrode design using readily available parts, yet can be easily scaled to any size trap with similar electrode spacings. We also derive the conditions for a crystal structure transition for two ions in the compact halo trap, the first nontrivial transition for Coulomb crystals in this geometry.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Evaluation of Graph Analytics Frameworks Using the GAP Benchmark Suite

The analysis of connected data is an increasingly important application in high-performance computing. Such analyses can reveal fraudulent patterns in financial transactions, optimize telecommunications networks, predict information flow in social networks, etc. However, the landscape of graph analytics is highly diverse. Graph algorithms stress processor architectures differently, and no one graph can represent all topologies. Consequently, no single approach or framework is expected to be optimal for all graph analytics problems. To help make sense of this diverse landscape, we evaluated four approaches to graph analytics: GraphBLAS, Galois, BGL17, GraphIt; and compare them against hand-tuned implementations that take advantage of hardware features on our test platform. Graph- BLAS formulates graph analytics as sparse linear algebra. Galois provides syntactic constructs for data parallelism over irregular data structures. BGL17 is a generic C++ template library for implementing graph algorithms. GraphIt provides a domain- specific language to describe and optimize graph algorithms. We use the GAP Benchmark Suite to establish baseline performance and guide the side-by-side evaluation of each framework. GAP consists of 30 tests: six graph analytics algorithms (breadth- first search, single-source shortest path, PageRank, betweenness centrality, connected components, and triangle counting) run on five graphs, each with different topological characteristics (e.g., high diameter, skewed degree distribution, high average degree). High-performance reference implementations are included for each benchmark algorithm. Because a graph can be loaded into memory a number of ways (e.g., flat file on disk, compressed sparse format, data frames, retrieved from SQL or NoSQL databases), our evaluation focused on computational performance rather than I/O. Our results show the relative strengths of each framework.

Graph algorithms, Benchmarking, shared-memory prog↗

Toward Optimal Transport Networks

Strictly evolutionary approaches to improving the air transport system a highly complex network of interacting systems no longer suffice in the face of demand that is projected to double or triple in the near future. Thus evolutionary approaches should be augmented with active design methods. The ability to actively design, optimize and control a system presupposes the existence of predictive modeling and reasonably well-defined functional dependences between the controllable variables of the system and objective and constraint functions for optimization. Following recent advances in the studies of the effects of network topology structure on dynamics, we investigate the performance of dynamic processes on transport networks as a function of the first nontrivial eigenvalue of the network's Laplacian, which, in turn, is a function of the network s connectivity and modularity. The last two characteristics can be controlled and tuned via optimization. We consider design optimization problem formulations. We have developed a flexible simulation of network topology coupled with flows on the network for use as a platform for computational experiments.

Alexandrov, Natalia↗

Towards string order melting of spin-1 particle chains in superconducting transmons using optimal control

Utilizing optimal control to simulate a model Hamiltonian is an emerging strategy that leverages the intrinsic physics of a device with digital quantum simulation methods. Here we evaluate optimal control for probing the nonequilibrium properties of symmetry-protected topological (SPT) states simulated with superconducting hardware. Assuming a tunable transmon architecture, we cast the evolution of these SPT states as a series of one- and two-site pulse optimization problems that are solved in the presence of leakage constraints. From the generated pulses, we classically simulate the time-dependent melting of the perturbed SPT string order across a six-site model with an average state infidelity of 10 -3 . The feasibility of these pulses as well as their efficient application indicate that high-fidelity simulations of string order melting are within reach of current quantum computing systems.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Optimal deployment of resources for maximizing impact in spreading processes

Significance Spreading processes play an increasingly important role in marketing, opinion setting, and epidemic modeling. Most existing algorithms for optimal resource allocation in spreading processes are based on topological characteristics of the underlying network and aim to maximize impact at infinite time. Clearly, realistic and efficient real-time allocation policies should consider both network properties and details of the dynamics; additionally, control may be applied only to a restricted set of accessible nodes, and impact should be maximized in a limited time window. We introduce a probabilistic targeting framework that incorporates the dynamics and encompasses previously considered optimization formulations. It is based on a scalable dynamic message-passing approach that allows for the solution of large real-world network instances.

60 APPLIED LIFE SCIENCES↗